Beta Engine: R^3 [Updated 6/5/14]

[YoshL]

Re: Beta Engine: R^3 [Updated 6/18/13]

I would go for average rank 1 in a month if rates on files (lets say below VC files) counted.

stuff like this. Encouraging players to get back into the game is bad?
I don't see it personally

[HalfStep]

Re: Beta Engine: R^3 [Updated 6/18/13]

Making it a difficulty threshold instead of a minimum rate threshold could also be a solution to bsing patterns on higher rates.

[foxfire667]

Re: Beta Engine: R^3 [Updated 6/18/13]

Eh, it's still probably not going to happen unless the community clamours hard for it though - I'd imagine it will be difficult changing the minds of many older players :P

Alright guys, let's all make a big noise about it and change the minds of the opposing community! Though majority rules probably really shouldn't influence something like this, unless legitimate reasons not sprung from bias or "because it's always been like that" are brought to the table (which really hammer terrible outcomes of adding 1.0+ rate scoring).

So far, some points brought up have been valid to raise, but giving them more thought seems to iron them out a bit.

- Though playing a song on a faster rate could potentially allow for certain patterns to be "cheated" easier than they are currently, it comes at the serious cost of trying to score well on the rest of the file. Considering that playing a difficult song on even 1.1/1.2x can vastly change the overall difficulty of the file (and even make those rolls too fast to hit) it evens out quite well. If someone AAA'd Skeletor on 1.3x, I think I would be happy to allow that to represent their 1.0 AAA.

- Forcing boredom as a game mechanic seems silly and discouraging for players actually trying to advance in the game.

- Mashing songs at a faster rate, as YoshL mentions, still means you are putting in the effort to hit the notes. Though it does raise the question of people exploiting it with special input sending keyboards or the usage or double setup, but that can be dealt with under the terms of rule breaking.

[TC_Halogen]

Re: Beta Engine: R^3 [Updated 6/18/13]

arc: for the sake of figuring out how to scale the windows, what would be the equivalent of default values?

[HalfStep]

Re: Beta Engine: R^3 [Updated 6/18/13]

Pretty sure default is supposed to be this
-116,5:-83,25:-50,50:-16,100:16,50:50,25:83,25:116,0

[Velocity]

Re: Beta Engine: R^3 [Updated 6/18/13]

judgeSettings = [
{t:-118, s:5, f:-3},
{t:-84, s:25, f:-2},
{t:-50, s:50, f:-1},
{t:-17, s:100, f:0},
{t:17, s:50, f:1},
{t:50, s:25, f:2},
{t:84, s:25, f:3}
];

-118:5,-84:25,-50:50,-17:100,17:50,50:25,84:25

[HalfStep]

Re: Beta Engine: R^3 [Updated 6/18/13]

Was a few ms off, my bad ^^

[arcnmx]

Re: Beta Engine: R^3 [Updated 6/18/13]

cmon i changed AJ's mind (i think) it's possible

Heh, true enough! Maybe it'll be easier than I thought.

Making it a difficulty threshold instead of a minimum rate threshold could also be a solution to bsing patterns on higher rates.

Eh, making a special case by difficulty is using it for the single purpose of grinding through levelranks faster... If you concede that rates mean you can bs patterns and shouldn't be allowed for the "hard" songs then this is just saying we don't care what happens to the easier leaderboards. It also obviously doesn't address the complaints that rates don't count against GTS, PA/bestscores, etc.

Alright guys, let's all make a big noise about it and change the minds of the opposing community! Though majority rules probably really shouldn't influence something like this, unless legitimate reasons not sprung from bias or "because it's always been like that" are brought to the table (which really hammer terrible outcomes of adding 1.0+ rate scoring).

So far, some points brought up have been valid to raise, but giving them more thought seems to iron them out a bit.

- Though playing a song on a faster rate could potentially allow for certain patterns to be "cheated" easier than they are currently, it comes at the serious cost of trying to score well on the rest of the file. Considering that playing a difficult song on even 1.1/1.2x can vastly change the overall difficulty of the file (and even make those rolls too fast to hit) it evens out quite well. If someone AAA'd Skeletor on 1.3x, I think I would be happy to allow that to represent their 1.0 AAA.

- Forcing boredom as a game mechanic seems silly and discouraging for players actually trying to advance in the game.

- Mashing songs at a faster rate, as YoshL mentions, still means you are putting in the effort to hit the notes. Though it does raise the question of people exploiting it with special input sending keyboards or the usage or double setup, but that can be dealt with under the terms of rule breaking.

I suppose it's less about majority rules and more about convincing the people who have a say... Admins? Game Managers? Iunno. It has my support though :P

As for using rates for mashing... Yes, you can make marginal gains by rate'ing VS Boss Battle, I don't believe that's a huge problem though. Cheating with double setups and such seems like a nonissue, it's not like FFR doesn't already have fast mashable songs such as RATO anyway.

Was a few ms off, my bad ^^

Yeah, just some rounding issues (also the windows technically start one ms after the value given). Looking at it again the defaults should perhaps be changed to be more like... -118, -84, -49, -18, 17, 50, 83
Generally it's calculated by {-3.5,-2.5,-1.5,-0.5,0.5,etc.}*1000/30

[qqwref]

Re: Beta Engine: R^3 [Updated 6/18/13]

I'd like to point out something subtle about rate modding... remember how files are stored, not with accurate millisecond timings or whatever, but with timings in frames?

Suppose you have a ~180 bpm roll, which will have a whole bunch of 2-frame intervals. Now let's speed it up to ~225 bpm (so 1.25 rate), where theoretically it wouldn't have any 2-frame intervals. The problem is that the file was already converted at 180 bpm, and now instead of having each note in the roll at a nice 1-note interval, you will have something like this:
- 1-frame intervals in the original song convert to 0.80 frames, which means a 20% chance of 0 frame and a 80% change of 1 frame
- 2-frame intervals in the original song convert to 1.60 frames, which means a 40% chance of 1 frame and a 60% chance of 2 frame
See what I mean here? Every 2-frame interval in the original roll has a 60% chance of still being a 2-frame interval. Even at 1.5 rate, 2-frame intervals have a 33.3% chance of staying that way. So speeding up the file actually will not make the roll that much better; it might end up easily cheatable, but it might just end up with really awkward and uneven frame intervals. And of course the entire file will be faster anyway. I think if you can easily PA Skeletor at 1.2 or 1.3 rate you can probably get much better AAAs anyway, which means that even if you do get that AAA it won't be an unfairly good score for you.


I think I would support allowing ratemodded scores to record, as long as the rate is high enough. I think a minimum of something like 1.2 rate would be okay. Anything too close to 1.0 just gives people the opportunity to randomly move notes around, and obviously anything below 1 is off limits.

edit: and maybe we should force rates to be multiples of something like .05 or .1 for score recording purposes, since otherwise we might get people trying a whole bunch of random rates (e.g. 1.204, 1.206, 1.208) just to try to slightly mess with the frame patterns.

[arcnmx]

Re: Beta Engine: R^3 [Updated 6/18/13]

Right, I meant to touch on that in my last post but was sidetracked. Agreed, but I was going to suggest a minimum of 1.1

[TC_Halogen]

Re: Beta Engine: R^3 [Updated 6/18/13]

edit: and maybe we should force rates to be multiples of something like .05 or .1 for score recording purposes, since otherwise we might get people trying a whole bunch of random rates (e.g. 1.204, 1.206, 1.208) just to try to slightly mess with the frame patterns.

Very good catch/idea.

[PriestREA]

Re: Beta Engine: R^3 [Updated 6/18/13]

I don't know if I've missed a memo or something, but I seem to always have to set my global offset to "3" before the notes are synced up with the song. At home I can play with a judge offset at 0, but at school I always need a judge of 1. As a result, I am substantially less accurate on R^3 than any other engine. I must have not read all the tips properly, because everyone else seems to have no problems.

I have tried it on Flash 9, 10, 11 and in-browser, cleared cache etc.

[arcnmx]

Re: Beta Engine: R^3 [Updated 6/18/13]

Use the right-click-on-speedmod-line-on-results-screen sekrit to see what the game thinks your judge offset should be. Different setups/computers/etc. can feel different, and with the fps changes that can become more obvious and pronounced. Even on the same computer fps mode vs 30fps mode can feel different and need specific offsets.

Also flash on my Linux system is bugged and needs a global offset of 6 or 7 for audio sync, no idea if that applies to other systems too.

tl;dr offsets until it feels comfortable.

[PriestREA]

Re: Beta Engine: R^3 [Updated 6/18/13]

Use the right-click-on-speedmod-line-on-results-screen sekrit to see what the game thinks your judge offset should be. Different setups/computers/etc. can feel different, and with the fps changes that can become more obvious and pronounced. Even on the same computer fps mode vs 30fps mode can feel different and need specific offsets.

Also flash on my Linux system is bugged and needs a global offset of 6 or 7 for audio sync, no idea if that applies to other systems too.

tl;dr offsets until it feels comfortable.

For me on both Mac and Win 8 I need a global offset of 3.
I will try the offset secret momentarily.

[Garquillex]

Re: Beta Engine: R^3 [Updated 6/18/13]

there should be a tournament to see who can AAA Excite Bike on the highest rate